CROI 2025 Abstract eBook

Abstract eBook

Poster Abstracts

463

WITHDRAWN

Results: The cumulative data indicated that LN DC frequencies correlated with plasma viral load (r=0.875; P <0.05; N=10). LN DCs from viremic individuals harbored an immunostimulatory profile with the ability to secrete pro inflammatory cytokines (IL-1β, IL-6), higher expression of co-stimulatory molecules (CD40, CD83, CD86) and were more potent stimulators of allogeneic CD4 T cell proliferation ex vivo (N=3; P <0.05) as compared to uninfected and ART-treated individuals, which was further increased upon TLR agonist stimulation. In-depth analyses of LN DC transcriptomic profile by scRNAseq further supported a specific and distinct transcriptomic profile of LN DCs during chronic infection associated with higher DC activation, Type I IFN and Type II IFN production, interferon-stimulated gene (ISG) signature, and activation of key STAT molecules (N=6). Interestingly, 18-26% of LN DCs from viremic individuals harbored detectable HIV transcripts. Notably, the comparison of HIV transcript+ versus HIV transcript- analyses further supported the enhanced expression of pathways involved in immune activation. While HIV transcripts were not detected after suppressive ART, LN DC gene expression profile still harbored markers associated with inflammation and immune activation (IL-1β signaling, TNFα-signaling, NF-κB activation), while several genes involved in antigen processing and presentation were downregulated. The involvement of epigenetic modifications in this specific profile remains to be determined. Conclusions: HIV-infected LN DCs harbor a specific gene signature associated with inflammation and immune activation, which is only partially restored after suppressive ART. These findings suggest that approaches targeting immune activation should consider the involvement of HIV-infected LN DCs. Mapping the Spatial Landscape of Lymph Nodes With Undetectable Versus Active HIV Replication Candace Liu 1 , Marta Calvet-Mirabent 2 , Michael Angelo 1 , Enrique Martin Gayo 3 1 Stanford University, Stanford, CA, USA, 2 Gladstone Institute, San Francisco, CA, USA, 3 Universidad Autónoma de Madrid, Madrid, Spain Background: Despite its effectiveness in suppressing active HIV replication, antiretroviral therapy (ART) does not eliminate the long-lived pool of HIV reservoir cells. Lymphoid tissues are a major reservoir even in the context of ART, but most studies on latent reservoirs focus on peripheral blood. The complex interplay between immune cells and the mechanisms that maintain viral persistence in lymphoid tissue remain incompletely understood. Advances in tissue imaging offer a unique opportunity to characterize the reservoir. Here, we used a spatial proteomic method to map immune microenvironments in HIV infected tissue. Methods: Lymph node (LN) tissue from 6 people with HIV (PWH; n=3 undetectable, n=3 viremic) were obtained from the Pathology Anatomy Unit from Hospital Universitario La Princesa. LNs from people without HIV (PWOH) were included as negative controls (n=5). We applied MIBI-TOF (Multiplexed Ion Beam Imaging by Time of Flight), to simultaneously image 42 proteins in these tissues. We profiled over 1.8 million cells and characterized the spatial relationships between 20 immune cell types in their native tissue context. The inclusion of the viral protein p24 allowed us to identify p24+ and p24- microenvironments and their functional signatures. Results: As expected, LNs from PWH had fewer CD4+ T cells compared to PWOH (p= 0.0006), while there were higher numbers of CD8+ T cells in both the whole tissue (p=0.024) and specifically within follicles (p=0.021). These CD8+ T cells expressed higher levels of Granzyme B and were enriched in p24+ LN (p=0.011). Further characterization of p24+ regions revealed higher levels of markers associated with activation and exhaustion of memory and effector CD8+ T cells (CD45RO, ICOS, TIGIT, Galectin9, TIM3), as well as inflammasome markers Caspase1 (p=0.03) and NLRP3 (p=0.3, FC=1.90). Using a new spatial niche detection method, we found that LN from PWH were enriched for CD8+ T cells near antigen-presenting cells (CD11c+, HLADR+), fibroblasts (SMA+), and endothelial cells (CD31+). Conclusions: Using spatial proteomics to map the immune landscape in LNs, we identified novel markers characterizing CD8+ T cell subsets enriched in PWH with different levels of viremia, implying that active viral replication confers long-term changes on the immune landscape in LNs. Collectively, these data provide new insights into complex cell networks associated with viral replication at an important tissue reservoir site, which could be relevant for future HIV cure strategies.

464

Distinct Signatures of Trained Immunity Reveal Heterogeneity Among Elite HIV Controllers Aarthi Talla 1 , Joao Lucas Lima Calandrini de Azevedo 2 , Michael Peluso 3 , Thomas Dalhuisen 3 , Rebecca Hoh 3 , Steven G. Deeks 3 , Sulggi Lee 3 , Ashish A. Sharma 2 , Rafick P. Sekaly 2 , for the Pathology Advanced Translational Research Unit (PATRU) 1 Immunasyst, Scottsdale, AZ, USA, 2 Emory University, Atlanta, GA, USA, 3 University of California San Francisco, San Francisco, CA, USA Background: Elite Controllers (ECs) are individuals who suppress HIV replication without treatment. Understanding these mechanisms can provide insights for developing novel therapeutic strategies. By learning from their responses, we can identify key pathways and immune functions that could enhance viral control in the HIV-infected population. This study aims to investigate ECs using a multiomics approach to understand how they achieve viral suppression. Methods: Samples from 13 ECs, 27 VCs, 10 Viremic Non-Controllers (VNCs), and 20 Healthy Donors (HDs) were collected. Plasma and PBMCs were analyzed for cytokine levels, plasma metabolites, and bulk RNAseq. A linear regression model was employed. Enrichment mapping identified transcriptomic modules, and K-means clustering unveiled heterogeneity among controllers. Bulk RNA modules were validated using an independent scRNAseq dataset and confirmed via in-silico deconvolution. Gradient boosting identified plasma cytokines and metabolites predictive of controller subgroups. Multiomics integration was performed using Diablo. Results: Transcriptomic analysis revealed 2 EC clusters and 3 VC clusters. CD4 counts and viral load within cohort between clusters (esp cluster 1 vs 2) are not confounding factors. Cluster 1 included ECs and VCs whose profiles resembled HDs (Figure 1A), enriched in classical monocytes with trained immunity, inflammasome and epigenetic remodeling signatures. These individuals showed equal or higher expression of these signatures compared to HDs. Cluster 2 comprised ECs and VCs with distinct profiles, with higher proportions of effector CD8 T cells with antiviral responses (Figure 1B). Cluster 3 included VCs and VNCs with higher expression of effector CD8 and antiviral responses. IFN-γ was the best predictor of the two EC classes, being lower in ECs with trained immunity and higher in those with effector CD8 (84% accuracy). Ceramides, primary bile acids, and fatty acids were metabolite predictors, with the first two higher in ECs with effector CD8 signatures, and fatty acids higher in ECs with trained immunity signatures (100% accuracy). Conclusions: The study reveals two distinct EC populations with differing immunological, metabolomic, and transcriptomic profiles. Recognizing this heterogeneity is crucial, as personalized therapeutic strategies could target specific pathways to enhance viral control. These findings suggest strategies to shift VCs and VNCs toward beneficial EC profiles, advancing efforts toward a functional HIV cure. The figure, table, or graphic for this abstract has been removed. Lymph Node Dendritic Cells Harbor an Altered Transcriptomic Profile Despite Years of Suppressive ART Riddhima Banga 1 , Erica Lana 1 , Francesco Procopio 1 , Mathilde Foglierini-Perez 1 , Matthias Cavassini 2 , Rafael Trunfio 1 , Sébastien Déglise 1 , Giuseppe Pantaleo 1 , Matthieu Perreau 1 , Andrea Mastrangelo 1 1 Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland, 2 Lausanne University Hospital, Lausanne, Switzerland Background: Despite significant advances in the efficacy of antiretroviral therapy (ART) and expanded access to people living with HIV, virologic suppression through ART alone is frequently insufficient to fully resolve the immune activation associated with HIV infection. Methods: We evaluated the influence of HIV infection on the functional and transcriptomic profile of lymph node (LN) myeloid dendritic cells (DCs) subsets during both chronic infection and after >5 years of suppressive ART by single cell RNAsequencing (scRNAseq).

Poster Abstracts

466

465

CROI 2025 115